Machine learning is increasingly engaged in a large number of important daily decisions and has great potential to reshape various sectors of our modern society. To fully realize this potential, it is important to und...
详细信息
Machine learning is increasingly engaged in a large number of important daily decisions and has great potential to reshape various sectors of our modern society. To fully realize this potential, it is important to understand the role that humans play in the design of machine learning algorithms and investigate the impacts of the algorithm on humans. Towards the understanding of such interactions between humans and algorithms, this dissertation takes a human-centric perspective and focuses on investigating the interplay between human behavior and algorithm design. Accounting for the roles of humans in algorithm design creates unique challenges. For example, humans might be strategic or exhibit behavioral biases when generating data or responding to algorithms, violating the standard independence assumption in algorithm design. How do we designalgorithms that take such human behavior into account? Moreover, humans possess various ethical values, e.g., humans want to be treated fairly and care about privacy. How do we designalgorithms that align with human values? My dissertation addresses these challenges by combining both theoretical and empirical approaches. From the theoretical perspective, we explore how to designalgorithms that account for human behavior and respect human values. In particular, we formulate models of human behavior in the data generation process and designalgorithms that can leverage data with human biases. Moreover, we investigate the long-term impacts of algorithm decisions and designalgorithms that mitigate the reinforcement of existing inequalities. From the empirical perspective, we have conducted behavioral experiments to understand human behavior in the context of data generation and information design. We have further developed more realistic human models based on empirical data and studied the algorithm design building on the updated behavior models.
We describe the operational algorithm being used to map global land cover on a quarterly basis using data from MODIS, This algorithm uses a supervised classification methodology and exploits a database of over 1000 tr...
详细信息
ISBN:
(纸本)0780363590
We describe the operational algorithm being used to map global land cover on a quarterly basis using data from MODIS, This algorithm uses a supervised classification methodology and exploits a database of over 1000 training sites identified from Landsat TM data globally. Land cover classes being mapped include the 17 categories defined by the International Geosphere-Biosphere program. However, the site database has been specifically designed to map a much richer set of classes based on the needs of the user community. The classification algorithms being developed for this purpose include decision trees and artificial neural networks. Recent results from prototyping efforts using AVHRR data demonstrate that the algorithms are performing well.
The speed and resource issues on algorithm design and implementation with hardware are discussed in this paper. There are two approaches to improve system-processing speed and to save logic resource have been proposed...
详细信息
ISBN:
(纸本)0780366778
The speed and resource issues on algorithm design and implementation with hardware are discussed in this paper. There are two approaches to improve system-processing speed and to save logic resource have been proposed, Furthermore quantitative analysis is performed to the results.
The paper shows summary of the author's research subjects from 1973 through 2012. Additional explanation on these subjects and related references are omitted because of space limitation. They will be given at pres...
详细信息
ISBN:
(纸本)9780769548937
The paper shows summary of the author's research subjects from 1973 through 2012. Additional explanation on these subjects and related references are omitted because of space limitation. They will be given at presentation.
The virtual experiment study is becoming a hot issue of network education study and attracts widespread attentions. This paper describes the important position of network virtual experiment system in the education fie...
详细信息
ISBN:
(纸本)9783642330292
The virtual experiment study is becoming a hot issue of network education study and attracts widespread attentions. This paper describes the important position of network virtual experiment system in the education field. It can conform to the requirements of remote education system and get rid of the restraints of traditional laboratory mode, so as to achieve the full sharing of teaching resources. Combined with the actual development work, this paper introduces the system architecture and drawing engine layer of virtual experiment system, the realization of the simulation frame layer and simulation realization layer. With the development of virtual reality technology and artificial intelligence, the virtual experiment system uses more and more applications of these advanced technologies to improve the user experience. And this is the major development trend of virtual experiment system.
For information consumers, it is very difficult to find the information they are interested in from the large amount of information; For information producers, it is also very difficult to make their own information s...
详细信息
For information consumers, it is very difficult to find the information they are interested in from the large amount of information; For information producers, it is also very difficult to make their own information stand out and attract the attention of the majority of users. In order to solve this contradiction, we built the de-correlation model based on principal component analysis, the prediction music track rating model, and the collaborative filtering recommendation model, so as to solve the problem that users can find high-quality music tracks from many music tracks. Through the analysis, it is concluded that the factors influencing users' evaluation of music tracks are the number of music track labels and indirect attention. Then, according to the linear regression theory, the model of predicting music track score is established, which can predict the user's score to music track. The mathematical model established in this paper has strong portability and can be extended to the fields of network, media, film and television.
Background. Immunological biomarkers have often been used as a complementary approach to support clinical diagnosis in several infectious diseases. The lack of commercially available laboratory tests for conclusive ea...
详细信息
Background. Immunological biomarkers have often been used as a complementary approach to support clinical diagnosis in several infectious diseases. The lack of commercially available laboratory tests for conclusive early diagnosis of leprosy has motivated the search for novel methods for accurate diagnosis. In the present study, we describe an integrated analysis of a cytokine release assay using a machine learning approach to create a decision tree algorithm. This algorithm was used to classify leprosy clinical forms and monitor household contacts. Methods. A model of Mycobacterium leprae antigen-specific in vitro assay with subsequent cytokine measurements by enzyme-linked immunosorbent assay was employed to measure the levels of tumor necrosis factor (TNF), interferon-gamma, interleukin 4, and interleukin 10 (IL-10) in culture supernatants of peripheral blood mononuclear cells from patients with leprosy, healthy controls, and household contacts. Receiver operating characteristic curve analysis was carried out to define each cytokine's global accuracy and performance indices to identify clinical subgroups. Results. Data demonstrated that TNF (control culture [CC]: AUC = 0.72;antigen-stimulated culture [Ml]: AUC = 0.80) and IL-10 (CC: AUC = 0.77;Ml: AUC = 0.71) were the most accurate biomarkers to classify subgroups of household contacts and patients with leprosy, respectively. Decision tree classifier algorithms for TNF analysis categorized subgroups of household contacts according to the operational classification with moderate accuracy (CC: 79% [48/61];Ml: 84% [51/61]). Additionally, IL-10 analysis categorized leprosy patients' subgroups with moderate accuracy (CC: 73% [22/30] and Ml: 70% [21/30]). Conclusions. Together, our findings demonstrated that a cytokine release assay is a promising method to complement clinical diagnosis, ultimately contributing to effective control of the disease.
According to the traditional application of the artificial projection method to escape wheel detection, this paper proposes a new method based on digital template, after the introduction of the working principle and a...
详细信息
ISBN:
(纸本)9783037856932
According to the traditional application of the artificial projection method to escape wheel detection, this paper proposes a new method based on digital template, after the introduction of the working principle and advantage, emphatically discuss the algorithm design. The experimental results demonstrate the proposed algorithm has high real-time ability, good reliability and more suitable for practical engineering application.
The study of algorithm design and control strategies based on mathematical models has become particularly important in the current context of increasing complex systems. The aim of this study is to explore how mathema...
详细信息
This paper discusses the algorithm design for a special type of structure optimization in the field of logic synthesis. The data structure used inhere is a kind of directed graph. We'll give the algorithm framewor...
详细信息
ISBN:
(纸本)7543909405
This paper discusses the algorithm design for a special type of structure optimization in the field of logic synthesis. The data structure used inhere is a kind of directed graph. We'll give the algorithm framework on the basis of analysis and comparison. A substantial logic network used as input data can demonstrate the efficiency and the feasibility of the software programmed in this algorithm.
暂无评论